Machine Learning Approaches for Predicting RNA-RNA/DNA Interactions

In this chapter, I introduce machine learning approaches for predicting RNA-RNA/DNA interactions, which are crucial for understanding noncoding RNA (ncRNA) functions. Advancements in deep learning techniques and the availability of large-scale interaction data from high-throughput sequencing methods have driven the development of these prediction tools. This review covers representative studies across different RNA families, including prokaryotic small RNAs (TargetRNA3), general RNA-RNA interactions (CheRRI), miRNAs (DeepMirTar), box C/D snoRNAs (snoGloBe), lncRNA-DNA triplexes (triplexFPP), and CRISPR guide RNA design (CRISOT). These machine learning-based methods often improve accuracy compared to traditional energy-based approaches. However, there are challenges such as the need for preventing overfitting and third-party validation. Future advancements are expected to enhance the generalization and applicability of these prediction tools, contributing to a deeper understanding of RNA functions.